|Abstract:||Silicon is the dominant material in the semiconductor industry, massively used in transistors, rectifiers, microchips, and other solid-state electronic devices. On top of that, its optoelectronic properties are largely utilized in photovoltaics, photodiodes, and photodetectors. However, in many optoelectronic devices silicon is limited by its band gap of 1.12 eV (_ = 1,110 nm) which precludes absorption of low energy infrared photons. For example, current photovoltaic devices suffer energy loss to a great extent due to non-absorption of a large portion of the incoming solar spectrum. Additionally, if the photoresponse of silicon could be extended to the infrared, silicon could be used to replace more expensive germanium based detectors.
My thesis research focuses on defect engineering of silicon to extend the photoresponse to the infrared regime. Recently, introducing defects at large concentrations several orders of magnitude larger than the equilibrium solubility limit has emerged as an approach to engineering the optical response of silicon. Specifically my thesis research addresses three aspects of hyperdoped semiconductors, which are as follows. (i) Carrying out first-principles investigations of gold hyperdoped silicon to provide detailed insights into recent extensive experimental investigations. This analysis helps to reveal the origins of the sub band gap optical response observed in this system, as well as to provide insights into the metastability and atomic scale mechanism of the loss of optical response upon thermal annealing. (ii) Moving beyond gold to consider alternative suitable transition metal dopants. It is of interest to identify candidate dopants that enhance optical absorption, and have sufficient solubility in silicon to remain dissolved in the semiconductor at high concentration and/or can become kinetically trapped to mitigate metastability. (iii) Using the insights gained from gold (Au) hyperdoped silicon systems, which have been extensively studied, to extend defect analysis towards gold hyperdoped germanium. This work involves carrying out high-accuracy first-principles modeling of candidate dopant defects in silicon, and assessing their promise for sub band gap absorption.
For gold hyperdoped silicon, first-principles density functional theory is used to first establish the origins of sub band gap optical absorption. While the experimentally synthesized systems likely contain a distribution of gold-related defects, experiments and the electronic structure analysis presented here strongly suggest that substitutional gold is responsible for the optical response. Unfortunately, the experimentally realized system optically deactivates upon thermal treatment. To understand the deactivation, I propose a mechanism for the evolution of atomic structure during thermal relaxation using simulations of energy barrier calculations, diffusion, and defect reactions. The dissociative mechanism, in which diffusion occurs by exchange of substitutional and interstitial sites via vacancy formation and annihilation, is identified as the likely path for deactivation. Further, the experimentally observed lattice contraction is explained by the presence of vacancies and gold-vacancy complexes in excess concentrations. The presence of these vacancies is attributed to the minimization of the strain that arises from the large volume of gold incorporated into silicon.
Towards the goal of identifying candidate hyperdoped materials for advanced optoelectronic devices such as intermediate band photovoltaics, I have carried out a systematic study of transition metal dopants in silicon in search for better dopants. To find other possible candidate dopants that could exhibit infrared absorption and, possibly, better solubility in silicon, I have assessed alternative transition metals such as Mn, Fe, Co, Ni, Cu and Ag using the optical design rules learned from prior investigations of chalcogen hyperdoped silicon. As the gold doped silicon study shows that not only the optical absorption but the (meta)stability of the system is also important, I have included an analysis of solubility and diffusion in silicon. This investigation also provided an opportunity to reassess the properties of transition metal defects in silicon, which had been studied extensively several decades ago, now using modern first-principles methods. As part of this analysis I have summarized their magnetic and electronic structure using density of states, orbital hybridization analysis, and band diagrams. Finally, I have extended the analysis of gold hyperdoped silicon to gold hyperdoped germanium. The relationship between measured carrier decay dynamics and some possible defect configurations are proposed in conjunction with recent experimentally achieved high
concentrations of gold and demonstrated optical response in germanium.